When designing high-availability computing systems, a premium is placed on providing fault-recovery mechanisms that can quickly regain full system performance with minimal downtime. For cost reasons, additional hardware and software specifically needed to perform fault recovery tasks should be reduced to a bare minimum.